Airborne Molecular Contamination: Formation, Impact, Measurement and Removal of Nitrous Acid (HNO2) Jürgen M

Home , Nitrous acid

MICROCONTAMINATION CONTROL | APPLICATION NOTE Airborne Molecular Contamination: Formation, Impact, Measurement and Removal of Nitrous Acid (HNO2) Jürgen M. Lobert, Reena Srivastava, and Frank Belanger – AMC Business Unit, Entegris, Inc. Franklin, MA USA | [email protected]

ABSTRACT Acidic gases have been a concern since the inception of 193 nm — technology. Initially focused on strong acids that cause corrosion Airborne molecular contamination (AMC) is a significant contributor on equipment or hazing of optics through the formation of salts to the loss of yield in semiconductor processes.1-4 The impact of (reactions with ammonia, for example),2-4 more recent concerns weak acids has only been considered for process technologies of are about the weaker acidic compounds, such as formic and 5-8 acetic acids, which can cause process degradation by affecting 22 nm and below. One such weak acid is nitrous acid (HNO2 or HONO), which has no demonstrated direct impact on processes or resist8 and other coatings. equipment, but has nevertheless been a target for removal by AMC Although no strict definition exists, weak acids are considered filtration. HNO is commonly formed on all surfaces in all environ- 2 to be those with a pKa of 3.2 or higher (acid strength, lower ments from NO gas, one of the main oxides of nitrogen formed 2 number indicates stronger acid), which includes formic, acetic from combustion processes and ambient air photochemistry. and other organic acids, but also nitrous acid (HNO , pKa = 3.3) This study investigated the behavior of the NO /HNO system 2 X X and hydrofluoric acid (HF, pKa = 3.2). In contrast to its strongly around typical AMC filter adsorbents. We find that NO gas passes acidic counterpart nitric acid (HNO , pKa = -2), however, HNO is through AMC filters unchanged, whereas NO is converted mostly 3 2 2 not a stable compound and its formation can easily be reversed to NO, but also to HNO at the low ppb level, increasing AMC load 2 into its components. As a result, HNO is rarely considered in downstream of filters. Various adsorbents can capture HNO , but 2 2 atmospheric chemistry cycles and very few dedicated studies filter lifetimes are short due to the release of the volatile compound exist.9, 10 In contrast to a variety of other nitrogen compounds over time. The recommendation is to critically evaluate the impact like NO , HNO is also not usually found in significant amounts of HNO on processes and equipment and adjust AMC filtration x 2 2 in the combustion process of nitrogen containing fuels.11 needs accordingly.

NOMENCLATURE INTRODUCTION — — A common misnomer used in the semiconductor industry is Controlling Airborne Molecular Contamination (AMC) has become “NO ”. The atmospheric science community has referred to the an indispensable part for high yield operations in semiconductor X mix of NO and NO gases as NO for decades, and the authors processing. What started decades ago with concerns about a few 2 X propose to continue the exclusive use of that acronym for this base compounds, such as ammonia1, expanded to controlling most purpose. The scientific community also uses the acronym NO measurable gas-phase contaminants classified as acids, bases, y to refer to the mix of other reactive nitrogen oxides, such as and organic compounds, plus some unclassified compounds like PAN (peroxy acyl nitrate), N O , N O etc., all commonly found hydrogen sulfide, ozone etc., down to parts per trillion (ppt, 10-12 2 3 2 5 in the atmosphere. mole fraction) levels in order to achieve semiconductor features below 22 nm, and to prevent critical dimensions from being degraded, coatings from delaminating, or metal surfaces from corroding.5-8 However, there is no such acronym in atmospheric B. Formation of HNO2 on surfaces chemistry to describe the sum of acidic species HNO2 Typical concentrations of HNO2 that are found by this and HNO3, which is what the semiconductor industry laboratory in various semiconductor environments, incorrectly calls “NOX”. Common use also suppresses the minus signs in ionic forms, leading to further such as air handlers, cleanrooms, scanner, and track -9 – tools, are a few parts per billion (ppb, 10 mole fraction), confusion between HNO2 in water (NO2 , nitrite) and mostly much below 10. However, much of the found the gas NO2. Here are the most common forms of HNO can be a result of an artifact described in the oxidized nitrogen gases: 2 next section.

NO: nitrogen oxide; NO2: nitrogen dioxide What complicates the evaluation of the impact of HNO2

NOX: Σ NO+NO2 is its instability. It is mainly formed from NO and NO2, two reactive gases from combustion processes (car NO : NO, NO , PAN, reactive oxides Y Σ 2 exhausts, heating systems, industrial emissions, etc.). (“odd nitrogen”) The primary formation of HNO2 on surfaces mentioned above explains why high surface adsorbents, such as HNO2: nitrous acid, a weak acid activated carbons, may be effective in triggering the without known impact conversion of NO2 to HNO2. Such highly porous – Nitrite: NO 2 , the ionic form of HNO2 surfaces are found in AMC chemical filters, and they in water can accelerate that formation to be much faster than through these more common, but slow, gas-phase HNO3: nitric acid, a strongly corrosive acid reactions:

– Nitrate: NO 3 , the ionic form of HNO3 NO + NO2 HNO2 (1) in water

2NO2 +H2O HNO2 + HNO3 (2) HNOX: Σ HNO2+HNO3 (suggested acronym)

Any formation of HNO2, however, is easily reversible – Virtual NOX : HNO 2 artifact formed in and yields the originating gas NO2 or NO and •OH 15 water solutions radicals. High surface carbon filters, however, cannot be avoided for most applications, because they are If any acronym is needed, instead of explicitly stating needed to remove organic AMC from the airstreams, the individual acids, the authors propose to adopt HNO x something that is not easily or cheaply done with non- to indicate the acidic character and avoid confusion carbon based adsorbents. This formation pathway is with NO . X currently the main concern in the industry, as it makes

AMC filters appear to be “outgassing” HNO2 when SOURCES AND REACTIONS FOR HNO2 applied to environments with significant NOX challenge.

A. Atmospheric HNO2

C. Formation of HNO2 in air samplers

Atmospheric HNO2 forms primarily through heterogeneous reactions on surfaces (e.g., particles), as Another issue that exacerbates evaluating the impact witnessed by highest concentrations found during of HNO2 is its measurement. HNO2 is unstable in air, nighttime, when available airborne particle surfaces but also forms and decomposes in aqueous solution. are highest.11, 13 The compound is also destroyed on The compound is not available or stable in pressurized surfaces (equilibrium) and through photolysis, reactions gas tanks. induced by light, explaining its distinct diurnal cycle The industry and associated laboratories most com- when local sources are absent.9 In the gas phase, HNO 2 monly use impingers /bubblers, devices that contain is an unstable, short-lived intermediate between the deionized water through which the air of interest is stable forms of NO and HNO . Concentrations of X 3 drawn with the use of a pump to dissolve all soluble actual gas-phase HNO in the atmosphere are usually 2 contaminants (usually acids and bases) in that water at the ppt level and up to 1 ppb, hence, not of concern for later analysis by ion chromatography 14 (Figure 1). to semiconductor processing at this time.

2 range, and was found to be as high as 1000 ppb in the

vicinity of combustion sources, the amount of HNO2 measured in impinger samples can quickly approach several ppb, in turn triggering alarms for the level of acidic AMC. The cited application note mentions a potential workaround for the artifact, but it involves duplicate equipment and is typically not done by laboratories (Figure 2).

Taken together, atmospheric HNO2 is negligible in most

cases. The amounts of HNO2 produced in impinger/ bubbler sampling can significantly contribute to re-

ported values, but the HNO2 formed by high surface adsorbents may be the biggest contributor to the

measured HNO2 (or dissolved nitrite) concentration in semiconductor environments.

IMPACT OF HNO2 ON PROCESS AND EQUIPMENT —

Based on a lack of understanding the chemistry and instability of the compound, the main concern about Figure 1. Typical deployment of a wet impinger/bubbler for the HNO in semiconductor environments appears to be determination of acids and bases. 2 perceptual, as that any presence of nominally “acidic” The authors reported before about the formation compounds is generally undesired. However, there 15 artifact of HNO2 in water impingers/bubblers. The are many organic acids that should trigger similar – effect was namedVirtual NOx (note the minus sign concerns, but are typically ignored, such as lactic indicating the ionic form), because the resulting HNO2 acid, which often is detected in semiconductor that forms in the aqueous solution is artificial, creating environment air samples. a ‘virtual’ and inflated signal for actual nitrite concentration. This effect is based on the conversion of NO Whereas strong acids can significantly impact process – steps and equipment at the ppb level, however, there and mostly NO2 to HNO2 (as nitrite, NO2 ) in water, is no published or direct impact of HNO on any semi- along the lines of Eq. 1, which happens faster in water 2 than in the gas-phase. conductor process step or equipment and in many cases, its existence at single digit ppb levels in semiconductor samples has been largely ignored for years by equipment and chipmakers.

The only issue the authors are aware of is that HNO2 may possibly facilitate corrosion of bare metals by

hydrogen sulfide (H2S), but it is the H2S causing the

corrosion, not the HNO2, the latter only attacking

a protective coating to let the H2S penetrate to the 17 metal. Note that HNO2 is only thought to facilitate, but not cause that impact. Removing the actual,

corrosive contaminant (H2S), which is easily achieved

with AMC filters, will alleviate that problem and HNO2 remains a compound without known, direct impact.

Figure 2. Concept of wet impingers configured for serial operation.

The authors found that about 1% of NOX in ambient air is converted to HNO2 in impingers. Given that typical atmospheric concentrations of NOX are in the 30-50 ppb

3 REMOVAL OF HNO2 AND OTHER AMC And all such solutions are either very expensive — compared to activated carbons, or require the use of electro-pneumatic systems (UV catalysts) that Generally, HNO2 can be removed from air streams like need power, maintenance, repairs, and can have strong acids, by adsorbing it with a base, and forming downtimes, something semiconductor applications a salt: do not tolerate. Acid (gas) + Base (gas or solid) Salt (solid) Most importantly, though, capturing or avoiding

In the most common case, a basic carbonate (MCO3, the formation of HNO2 in air handler systems, FFU with M being a metal) can be used either as a pure cleanroom protection or even at the tool level does granulate or, to improve capture efficiency and capacity, not eliminate the problem. NO2 gas is converted to as a coating on a porous carrier, such as carbon, zeolite, NO to some extent on AMC filters, but much of it or others. Carbonates are inorganic compounds, fairly penetrates the filters and makes it into the cleanroom unreactive, aside from their basicity, and do not volatilize, and into the process tools, due to its volatility and but they are weakly basic, hence, acid/base reaction is high concentrations in the atmosphere. We reported limited, so is the stability of the resulting salt. Hydrox- that a few percent of the ambient NOX gets converted ides such as NaOH or KOH can also be used and have to HNO2. much higher basicity (and toxicity!), but they react Any additional AMC filter or surface downstream of with atmospheric carbon dioxide (CO2) to form the the initial filtration step will again form HNO2 from the weaker carbonates Na2CO3 or K2CO3. remaining NO2. This problem will propagate all the

Another approach is to use a cationic ion exchanger, way inside process tools, where more HNO2 can be usually granular polymers that have their surface mod- formed on the wafer itself, or any other hard surface, ified to carry organic amines. The basicity of these can because surface-to-air volume ratios are higher indoors be strong, resulting in a stronger bond of weak acids than in the atmosphere.18 The issue can be minimized such as HNO2. with AMC filters, but not completely eliminated, because

complete filtration of ambient NO2 would require All other approaches to capturing acids are combina- massive amounts and frequent change of adsorbents. tions of these two basic principles and may involve fabrics or membranes or other carriers and chemicals. All of these approaches also work for strong acids, such as hydrochloric acid (HCl) and sulfuric acid (H2SO4). MEASUREMENT OF HNO2 FOR THIS STUDY —

Whereas it is possible to prevent the formation of HNO2 HNO can be measured with no-contact, long-range by using non-carbonaceous adsorbents for dedicated 2 optical methods such as DOAS,9 but their setup and applications of acid removal, the typical semiconductor operation is involved and unsuitable for small-scale environment requires to control all AMC, acids, bases, measurements that require close proximity to the and especially organic compounds. The only cost test site (e.g., AMC filter outlets). Denuder systems effective solution to removing organic AMC is through are another solution, and similar in functionality to the use of activated carbon adsorbents. impingers/bubblers, but more involved.9 IMS online There are other adsorbents for the removal of organics, monitors can detect acidic AMC, but it is unknown 19 such as ion exchange solutions and modified zeolites if they specifically react to the presence of HNO2. or the use of catalytic destruction, often aided by Chemiluminescence monitors are likely to detect the ultraviolet light, that all have been demonstrated to compound, but their converter system will convert work well for the removal of specific compounds, but it to NO, hence, they cannot detect the difference 9 none of these work well for the wide range of C3 to between NO2 and HNO2.

C26 contaminants that are typically monitored and of Contrary to what was speculated to be a workaround,15 concern to cause process issues. using dry adsorbent sample trap media (a base-coated solid state pad or adsorbent) was found to produce – nearly as much Virtual NOx as wet impingers, hence, did not provide any advantage for this study.

4 – 100 Based on the formation of Virtual NOx mentioned above, this study employed dual, serial impingers

(Figure 2), where the air enters the first and exits the 90 second impinger. The first impinger captures all – actual HNO2 in the inlet air, plus the Virtual NOx 80 produced in the impinger from ambient NOX. The

– Removal Eciency (%) second impinger only produces Virtual NOx , absent 70 of actual HNO2, which is soluble enough to be entirely scrubbed in the first device. The numerical difference 0 500000 1000000 1500000 Estimated Lifetime (ppb-h) between the impingers yields the actual HNO2. The Figure 3. Lifetime curve for HNO3 removal on activated carbon-based conversion of NOX to HNO2 is only 1-3% of the NO2 media. After initial 100% capture efficiency, HNO3 slowly breaks concentration, hence, the second impinger is exposed through. to nearly as much (99-97%) of the NO2 that the first – HNO is a strong acid (pKa = –2) and is efficiently impinger experiences, and Virtual NOx production 3 is about the same in both. removed with carbonaceous AMC filters, exhibiting long lifetimes / high capacities. Lifetime estimates For the setup of testing different adsorbents for the here are expressed in ppb-h (parts per billion - hours), removal of HNO2 from air streams, we employed a product of the challenge concentration and time several different test concepts. (Figure 3). This metric enables easy calculation of filter lifetime in hours by dividing the ppb-h metric • Deep beds: Many tests were carried out in deep bed by the measured, average concentration. stations, where a tubular device was filled with the adsorbent, much akin to a gas purifier. However, B. Measurement and Removal of HNO2 flow rate was adjusted to mimic the flow across an AMC filter in real world applications. This method Nitrous acid is a weak acid (pKa = 3.3) and often in

is best for comparing many different adsorbents. equilibrium with NOX or HNO3. As described, the low acidity affects how it is formed and detected and can • Patch tests: The next larger setup employed 50 mm create artifacts, artificially inflating actual HNO2 air diameter patches of AMC filter material, where the concentrations. In addition, HNO2 can form on surfaces, adsorbent is sandwiched between layers of fibrous including wafers, which is not detected by air samples. scrim and loft, as it is used in AMC filters. Removal of HNO undergoes a similar process as HNO , • AMC filters: To validate real world performance, we 2 3 with the added complication that additional HNO is carry out full size AMC filter tests, as they are deployed 2 formed from NO in air, which is ubiquitous and signifi- in semiconductor environments. X cant. Combustion processes in urban environments frequently cause a background signal of 30-100 ppb

of NOX, usually a third of which is NO2. Closer to

RESULTS FOR HNO2 REMOVAL AND PRODUCTION ON combustion sources, NO is more prevalent (it is the AMC FILTER ADSORBENTS original nitrogen oxide formed in the process), further — away from combustion sources, NO gets increasingly

oxidized to NO2. The latter is mostly responsible for A. Measurement and Removal of HNO3 the heterogeneous formation of HNO2. This adds to

Nitric acid (HNO3) is well captured by impingers and the overall acid load and diminishes the adsorbent – dissolves as nitrate (NO3 ). It is stable in solution and lifetime by exhausting the capacity faster. strongly corrosive to equipment and materials. Easy to calibrate and handle, measurement detection limits are a few ppt.

5 C. Conversion of NO2 on Carbonaceous Adsorbents D. Formation of HNOX on Carbonaceous Adsorbents

The conversion of NOX can be demonstrated on In the process of NOX conversion, HNO2 is formed common carbon adsorbents. Whereas NO passes from both NO and NO2, whereas HNO3 is only formed through the filter almost unchanged (Figure 4), NO2 from NO2, but was not detected downstream of NO. does not, but is rather retained and slowly breaks Formation of HNO from NO was low, about 1% of the through to about 20% of the original challenge 2 NO challenge. Formation from NO2, however, was (Figure 5), while the other 80% of the NO2 is very significant, but delayed, as the formed HNO is converted to NO (Figure 6). 2 adsorbed until the adsorbent capacity is exhausted 1200 (Figure 7). We found that about 5-10% of the NO2

1000 gets converted to HNO2, substantially more than we found happening in water impingers reported above. 800

NO2 conversion to HNO3 was also found, but only 600 about 1% of the NO2 challenge was detected as 400 HNO3. HNO3 is a strong acid and is retained well Concentration (ppb) by carbonaceous adsorbents. 200

0 Based on these results, the best AMC filtration solution 0 500 1000 1500 2000 requires the use of a strong base coating or material Time (h) to maximize HNOX capacity and lifetime. The strength Figure 4. NO gas passes through various carbon-based carbon adsorbents of the base coatings, however, is often restricted by virtually unchanged. Challenge concentration was about 1 ppm (=1000 ppb), which is found unchanged downstream of the adsorbent. material processing limitations. We also found that coating an activated carbon with more than 15% 200 chemical reduces or eliminates its capacity for organic AMC. 150

70 100 60

50 50 Concentration (ppb)

40 0 0 100 200 300 400500 600700 30 Time (h) Concentration (ppb) 2 20 Figure 5. A challenge of 1 ppm NO2 on various carbon adsorbents with increasing base character shows initial retention (while also being HN O 10 converted to NO), then a gradual and accelerating breakthrough until it levels off. 0 0 100 200 300 400500 600 1000 Time (h)

Figure 7. Formation and breakthrough of HNO2 on various carbon-based 750 adsorbents over time. Increasing base strengths (▲< ♦< ■< ■) on the adsorbent delays breakthrough and increases capacity, as expected.

500

Concentration (ppb) 250

0 0 100 200 300 400500 600 Time (h)

Figure 6. Conversion of NO2 to NO. These are downstream concentrations of NO when the same adsorbents of Figure 5 were challenged with 1 ppm of

NO2. The early, gradual increase suggests an accelerating conversion, likely assisted by the increasingly acidic character of the adsorbent.

6 E. Alternative Adsorbents Finally, we want to note that some tests indicated the formation of formic acid at the ppt level on increasingly As mentioned, the use of ion exchange adsorbents acidic media (either loaded carbonaceous adsorbents with basic character allows to capture acidic com- or cationic ion exchangers) in the presence of carbon pounds without releasing them over time. We also or perhaps carbon dioxide. This has not been studied did not find any NO conversion on cation exchangers. 2 yet, but to fully understand the behavior of acidic There are, however, four concerns about cationic ion AMC around adsorptive media warrants further exchange resins. investigation. First and most important is their very strong release of amines (causing T-topping) and associated fish-like odor. Second, if all of the NO gets passed through 2 CONCLUSIONS the AMC filter stage unchanged, it is available, and — makes it more likely, to form HNO2 on any surface, all the way to the wafer. Third, these adsorbents are HNO2 is not a stable acid or AMC, it is formed and 5-10 times more expensive than carbon-based adsor- destroyed in-situ on surfaces. It is difficult to accurately bents. Lastly, this type of adsorbent is limited in applica- detect and quantify due to measurement artifacts. tions and unable to capture organic or base AMC. HNO2 formation happens in the presence of NO2 gas, a combustion product in the atmosphere, whereas If deployed in conjunction with carbon-based adsor- formation from NO gas is negligible. bents to capture organic AMC or to alleviate the

There is no known, direct impact of HNO2 on any amine odor, the problem of NOX conversion and semiconductor process or equipment. The concern HNOX formation returns. If the ion exchange is deployed in a separate, downstream layer, it may about HNO2 in semiconductor environments appears to be largely perceptual based on its nominally acidic delay the carbonaceous HNO2 formation, but at high cost, and the amine odor will be released into the character. environment unchecked. There are cost-effective solutions to delay the formation

Non-carbon based adsorbents carrying some form and release of HNO2, but lifetime/capacity of such AMC filter solutions is limited. Some solutions to prevent its of permanganate (usually KMnO4 or NaMnO4) are formation exist, but they are either expensive, or outgas available to convert (but not remove) NO2, but were amines, and most do not capture any other AMC type. also found to produce HNO2 at similar levels compared to carbonaceous adsorbents, albeit with a more Good AMC filter solutions require a balance of chemical shallow release curve. In addition, permanganate performance (removal efficiency, capacity), physical does not capture organic or base AMC, hence, has performance (pressure drop, size, weight,…), operational very limited applicability. Being a strong oxidant, it may concerns (toxicity, smell, outgassing), and cost. Most also trigger unwanted conversions of organic AMC to importantly, AMC filter solutions need to work for volatile species that do not have long lifetimes on filters. all environments and target a broad range of AMC Similar limitations apply to using titanium oxides (TiO ), 2 types, preventing the sole use of specialized solutions. which were reported to destroy NO and HNO .20 2 2 Characterizing the AMC spectrum through competent Other types of adsorbents without carbon content analysis and customizing the filtration solutions to are molecular sieves (zeolites) or similar minerals, address all above concerns is important. also organic ion exchangers without acidic or basic To minimize cost, effort, and risk in broad-band AMC character, where pore size can be tailored to capture protection, the recommendation is to critically evaluate, specific AMC types, but none of these are broad-band quantify, and publish any actual impact of HNO or adsorbents and they remove only a few, targeted 2 reduce emphasis on its capture, if no impact is found. compounds.

The potential exists to use porous, non-carbon adsorbents coated with a base material to capture acids. Some test were carried out in this laboratory, but we found that capture efficiency and capacity are much reduced compared to carbon-based adsorbents.

7 References

1 11 Devon Kinkead and Monique Ercken., “Progress in Qualifying and Andy R. Reisinger, “Observations of HNO2 in the Polluted Winter Quantifying the Airborne Base Sensitivity of Modern Chemically Atmosphere: Possible Heterogeneous Production on Aerosols”, Amplified DUV Photoresists,” Proceedings, Vol 3999, Advances in Atmospheric Environment, vol. 34-23, pp. 3865-3874, 2000. Resist Technology and Processing XVII; https://doi.org/10.1117/ https://doi.org/10.1016/S1352-2310(00)00179-5. 12.388362, 2000. 12 Jürgen M. Lobert, Dieter H. Scharffe, Wei-Min Hao, Thomas A. 2 John K. Higley and Michael A. Joffe, “Airborne Molecular Kuhlbusch, Ralph Seuwen, Peter Warneck, and Paul J. Crutzen. Contamination: Cleanroom Control Strategies”, Solid State "Experimental Evaluation of Biomass Burning Emissions: Nitrogen Technology, pp. 211, July 1996. and Carbon Containing Compounds". Global Biomass Burning: Atmospheric, Climatic and Biospheric Implications, edited by 3 Mats Ekberg et al., “Airborne Molecular Contamination Control in J. S. Levine, pp. 289-304, MIT Press, Cambridge, Mass., 1991. the Micromirror SLM-Based Deep Ultraviolet DUV SIGMA7300 Laser Pattern Generator”, Proc. SPIE, vol. 5377, Optical Microlithography 13 Luke D. Ziemba, “Heterogeneous Conversion of Nitric Acid to XVII, (28 May 2004); https://doi.org/10.1117/12.535833. Nitrous Acid on the Surface of Primary Organic Aerosol in an Urban Atmosphere”, Atmospheric Environment, vol. 44-33, pp. 4081-4089, 4 Jitze Stienstra, Joseph Wildgoose, Jürgen Lobert, Christpher Vroman, 2010; https://doi.org/10.1016/j.atmosenv.2008.12.024 and David Ruede, “A 21st Century Approach to Airborne Molecular Contamination Control”, Gases & Instrumentation, vol. 4-2, pp. 12-15, 14 Methods of Air Sampling and Analysis, Intersociety Committee March 2010. TD890.M488, ed. James P. Lodge Jr., Third Edition, Lewis Publishers, 1989. 5 Paola González-Aguirre, Hervé Fontaine, Carlos Beitia, Jim Ohlsen, 15 – and Jorgen Lundgren, “Control of HF Volatile Contamination in Jürgen Lobert, Anatoly Grayfer, and Oleg Kishkovich, “Virtual NOX , FOUP Environment by Advanced Polymers and Clean Gas Purge”, a Measurement Artifact in Wet Impinger Air Sampling”, Entegris Solid State Phenomena, ISSN: 1662-9779, vol. 219, pp. 247-250 application note APN000015, Sep 2006. Available online for public https://doi.org/10.4028/www.scientific.net/SSP.219.247, 2015. download.

6 T.Q. Nguyen, H. Fontaine, Y. Borde, and V. Jacob. “Identification and 16 Tyler M. Moulton, Emily C. Zaloga, Katherine M. Chase and Jürgen Quantification of FOUP Molecular Contaminants Inducing Defects M. Lobert. “An Analytical Method for the Measurement of Trace in Integrated Circuits Manufacturing”, Microelectronic Engineering, Level Acidic and Basic AMC Using Liquid-Free Sample Traps”, Proc. vol. 105, pp. 124-129, 2013. SPIE, vol. 9050, Metrology, Inspection, and Process Control for Microlithography XXVIII, 90502B, April 2, 2014; https://doi.org/ 7 P. Gonzalez-Aguirre, H. Fontaine, C. Beitia, J. Ohlsen, and J. 10.1117/12.2046190. Lundgren, "FOUPs Against AMCs: The HF Case, Future Fabs International", vol. 42, pp. 80-86, 2012. 17 Personal communication, unpublished.

8 Nicolas Pic et al., “Defectivity Decrease in the Photolithography 18 Sasho Gligorovski. “Nitrous Acid (HONO): An Emerging Indoor Process by AMC Level Reduction Through Implementation of Novel Pollutant”, J. Photochem. Photobio. A: Chemistry, vol. 314, Filtration and Monitoring Solutions”, Proc. SPIE, vol. 7638, Metrology, pp. 1-5 (2016). Inspection, and Process Control for Microlithography XXIV, 76380M (2 April 2010); https://doi.org/10.1117/12.845591. 19 Personal communication, unpublished.

9 Gerhard Lammel and J. Neil Cape, “Nitrous Acid and Nitrite in the 20 Adrien Gandolfo et al., “The Effectiveness of Indoor Photocatalytic

Atmosphere”, Chem, Soc. Rev., vol. 25, pp. 361-369, 1996. Paints on NOX and HONO Levels”, Appl. Catalysis B: Environmental, vol. 166-167, pp. 84-90 (2015). 10 Peter Warneck, “Nitrogen Compounds in the Troposphere” in: Chemistry of the Natural Atmosphere, International Geophysics Series Volume 71, Academic Press, Inc., eds Renata Dmowska and James R Holten, 1999. This application note was originally published by Jürgen M. Lobert, Reena Srivastava, and Frank Belanger, "Airborne Molecular Contamination:

Formation, Impact, Measurement and Removal of Nitrous Acid (HNO2)", ASMC Proceedings, ISBN 978-1-5386-3748-7, 180-185, 2018.

FOR MORE INFORMATION Please call your Regional Customer Service Center today to learn what Entegris can do for you. Visit entegris.com and select the Contact Us link to find the customer service center nearest you.

Corporate Headquarters Customer Service 129 Concord Road Tel +1 952 556 4181 Billerica, MA 01821 Fax +1 952 556 8022 USA Toll Free 800 394 4083

Entegris®, the Entegris Rings Design®, and other product names are trademarks of Entegris, Inc. as listed on entegris.com/trademarks. All third-party product names, logos, and company names are trademarks or registered trademarks of their respective owners. Use of them does not imply any affiliation, sponsorship, or endorsement by the trademark owner. ©2018 Entegris, Inc. | All rights reserved. | Printed in the USA | 4511-10162ENT-1018